Age | Commit message (Collapse) | Author |
|
|
|
The r300 compiler can now emit instructions that select from the presubtract
source. A peephole optimization has been added to convert instructions like:
ADD Temp[0].x, none.1, -Temp[1].x into the INV (1 - src0) presubtract
operation.
|
|
Those are:
- dead-code elimination
- constant folding
- peephole (mainly copy propagation)
- register allocation
There are some bugs which I need to track down.
Also fix up the descriptions of all the debug options.
|
|
|
|
|
|
First list compiler passes in an array, then run the new function rc_run_compiler.
Every backend may need a different set of passes.
This cleans up the mess in r3xx_compile_vertex_program.
|
|
|
|
|
|
Wine likes to create a *lot* of constants, exceeding the size of the constant
file in hw.
|
|
|
|
|
|
Single loops work, but nested loops do not.
|
|
Needed for vertex shaders too.
|
|
This saves constant register space for r300g, which doesn't need
this feature.
|
|
|
|
This maybe breaks the vert compiler. Hopefully not.
|
|
Signed-off-by: Alan Coopersmith <alan.coopersmith@sun.com>
Reviewed-by: Corbin Simpson <MostAwesomeDude@gmail.com>
|
|
|
|
This replaces the old NQSSADCE code with the same functionality, but quite
different design. Instead of doing a single integerated pass, we now build
explicit data structures representing the dataflow.
This will enable analysis of flow control instruction, and could potentially
open an avenue for several dataflow based optimizations, such as peephole
optimization, fusing MUL+ADD to MAD, and so on.
|
|
In particular, this removes the dependency on prog_instruction, which
unfortunately creates some code duplication, but also opens a path towards
adding some hardware-specific things in there.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
This really doesn't belong into the compiler itself, since the compiler
should eventually be independent of Mesa's program representation.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
In the long run, it's probably better to just get rid of InputsRead and
OutputsWritten.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Attribute indices will probably be different in Gallium, so make the compiler
independent of magic values.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Adding constants is used in a number of non-native instruction
rewrites, and it required us to keep copies of modified gl_programs
around. This is a first step towards ending this.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Unfortunately, this does cause some code duplication (which we can hopefully
eliminate eventually).
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
This is just the first step of refactoring. The separation is not yet
clean enough with this commit.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
The goal is to convert both Mesa and TGSI programs into an intermediate format
that happens to be convenient for us.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|
|
This is in preparation of sharing the fragment program compiler with Gallium:
Compiler code is moved into its own directory and modified so that it no
longer depends on driver structures.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
|